
********************************************************************************************************************
*** Pro-equality initiatives increase expressed sexism among men but may improve trust among women football fans ***
********************************************************************************************************************

*** Authors:  
*** Victor Araújo (University of Reading) 
*** Malu A. C. Gatto (University College London)

*** Date: 02 June 2025


*** IMPORTANY NOTE: 
*** This do-file carries out the analyses conducted with observational data from our national survey with Brazilian respondents (reported in Figure 1 in the main manuscript). 
*** To replicate the analyses conducted with experimental survey data (reported Figures 3, 4, and 5 in the main manuscript), please see "AraujoGatto_Experimentalalanalysis.do". 



*** Installing required packages
ssc install estout, replace
ssc install blindschemes, replace
set scheme plotplain
ssc install coefplot, replace
ssc install cibar, replace


************************
*** Opening the data ***
************************

clear 
import delimited 
* Add your path to "AraujoGatto_Observationaldata.csv" 


****************************
*** Generating variables ***
****************************

****** Socio-demographic characteristics of respondents

tab q4
gen women = q4
recode women (1=0) (2=1) (8888 9999 =.)
tab women

tab q9
gen pretopardo = q9 
recode pretopardo (2 4 = 1) (1 3 5 6 =0) (8888 9999 =.) 
tab pretopardo

tab q6
gen age = q6
recode age (9999 =.)

tab q10
gen educ = q10 
recode q10 (9999=.)

tab q11
gen income = q11
recode income (8888=.) (9999=.)



label variable women "Woman" 
label variable age "Age" 
label variable pretopardo "Black" 
label variable educ "Education" 
label variable income "Income"


****** Identification with/importance of collective sources of social identity 

tab q21
gen tem_rel = 1
replace tem_rel = 0 if q21==11
replace tem_rel = 0 if q21==8888
replace tem_rel = 0 if q21==9999
tab tem_rel


destring q22, gen(importancia_rel) force
tab importancia_rel
recode importancia_rel (1 8888 9999 . = 0) (2 3 = 1) 
tab importancia_rel


tab q23
gen tem_part = 1
replace tem_part = 0 if q23==15
replace tem_part = 0 if q23==8888
replace tem_part = 0 if q23==9999
tab tem_part


destring q24, gen(importancia_part) force
tab importancia_part
recode importancia_part (1 8888 9999 . = 0) (2 3 = 1) 
tab importancia_part


tab q25
gen tem_time = 1
replace tem_time = 0 if q25==14
replace tem_time = 0 if q25==8888
replace tem_time = 0 if q25==9999
tab tem_time


destring q26, gen(importancia_time) force
tab importancia_time
recode importancia_time (1 8888 9999 . = 0) (2 3 = 1) 
tab importancia_time



tab q27
gen tem_mov = 1
replace tem_mov = 0 if q27==7
replace tem_mov = 0 if q27==8888
replace tem_mov = 0 if q27==9999
tab tem_mov


destring q28, gen(importancia_mov) force
tab importancia_mov
recode importancia_mov (1 8888 9999 . = 0) (2 3 = 1) 
tab importancia_mov


****** Left-right ideology

destring q42, gen(leftright) force
tab leftright



*****************************************************************************
******* Figure 1: Identification with and importance of football team *******
*****************************************************************************


graph bar (mean) tem_part (mean) tem_mov (mean) tem_rel (mean) tem_time, ylabel(0(.2)1) title("(A) Identification") blabel(bar) legend(cols(4)order(1 "Political party" 2 "Social movement" 3 "Religion" 4 "Football team")) 
graph save tem, replace

graph bar (mean) importancia_part (mean) importancia_mov (mean) importancia_rel (mean) importancia_time,  ylabel(0(.2)1) title("(B) Importance") blabel(bar) legend(order(1 "Political party" 2 "Social movement" 3 "Religion" 4 "Football team"))
graph save importancia, replace

grc1leg tem.gph importancia.gph, legendfrom(tem.gph)




****************************************
*************** APPENDIX ***************
****************************************


**********************************************************************
*** Appendix A: Football teams as non-value-oriented organisations ***
**********************************************************************

*** Generate variables to identify supports of each of Brazil's top 10 largest football clubs
*** Source of information on support base: https://ge.globo.com/futebol/noticia/2023/04/25/maiores-torcidas-do-brasil-pesquisa-atlas-mostra-flamengo-corinthians-e-sao-paulo-no-top-3.ghtml


gen america =. 
replace america = 1 if q25==10000 

gen athletico =. 
replace athletico = 1 if q25==10001

gen atletico =. 
replace atletico = 1 if q25==1

gen bahia =. 
replace bahia = 1 if q25== 10002

gen botafogo =. 
replace botafogo = 1 if q25== 2

gen ceara =. 
replace ceara = 1 if q25==10003

gen corinthians =. 
replace corinthians = 1 if q25==3

gen coritiba =. 
replace coritiba = 1 if q25==10004

gen cruzeiro =. 
replace  cruzeiro = 1 if q25==4

gen flamengo =. 
replace flamengo = 1 if q25==5

gen fluminense =. 
replace fluminense = 1 if q25==6

gen fortaleza =. 
replace fortaleza = 1 if q25==10005

gen gremio =. 
replace  gremio = 1 if q25==7

gen internacional =. 
replace  internacional = 1 if q25==8

gen nautico =. 
replace nautico = 1 if q25==10006

gen palmeiras =. 
replace  palmeiras = 1 if q25==9

gen paysandu =. 
replace  paysandu = 1 if q25==10007

gen remo =. 
replace  remo = 1 if q25==10008

gen santos =. 
replace santos = 1 if q25==10

gen santacruz =. 
replace santacruz = 1 if q25==10009

gen saopaulo =. 
replace  saopaulo = 1 if q25==11

gen sport =. 
replace sport = 1 if q25==10010

gen vasco =. 
replace vasco = 1 if q25==12

gen vitoria =. 
replace vitoria = 1 if q25==10011


*** Figure


twoway ///
(kdensity leftright if flamengo==1, lcolor(red) lpattern(shortdash)) ///
(kdensity leftright if corinthians==1, lcolor(black) lpattern(shortdash)) ///
(kdensity leftright if saopaulo==1, lcolor(orange) lpattern(shortdash)) ///
(kdensity leftright if palmeiras==1, lcolor(green) lpattern(shortdash)) ///
(kdensity leftright if vasco==1, lcolor(grey) lpattern(shortdash)) ///
(kdensity leftright if cruzeiro==1, lcolor(blue) lpattern(shortdash)) ///
(kdensity leftright if gremio==1, lcolor(eltblue) lpattern(shortdash)) ///
(kdensity leftright if atletico==1, lcolor(dimgray) lpattern(shortdash)) ///
(kdensity leftright if bahia==1, lcolor(midblue) lpattern(solid)) ///
(kdensity leftright if internacional==1, lcolor(maroon) lpattern(shortdash)) ///
, legend(label(1 "Flamengo") ///
         label(2 "Corinthias") ///
         label(3 "São Paulo") ///
         label(4 "Palmeiras") ///
         label(5 "Vasco") ///
         label(6 "Cruzeiro") ///
         label(7 "Grêmio") ///
         label(8 "Atlético") ///
         label(9 "Bahia") ///
         label(10 "Internacional")) ///
  title("") ///
  ylabel(, angle(horizontal)) ///
  legend(cols(1) position(3)) ///
  xtitle(Left-right ideology) ///
  ytitle("K-density (left-right ideology)")

 graph save clubideology, replace
 

**************************************************************
******* Appendix J: Profile of Brazilian football fans *******
**************************************************************


logistic tem_time women pretopardo age educ income, robust 
coefplot, drop(_cons) title("(A) Identification") xline(1, lcolor(red)) xlabel(-.4(.2)1.4) eform xtitle(Odds ratio)
graph save identification_reg, replace

logistic importancia_time women pretopardo age educ income, robust 
coefplot, drop(_cons) title("(B) Importance") xline(1, lcolor(red)) xlabel(-.4(.2)1.4) eform xtitle(Odds ratio)
graph save importance_reg, replace

*** Figure

graph combine "identification_reg" "importance_reg", title("") cols(2) 
graph save fansprofile, replace


